Is it possible to train a neural network with particle swarm optimization using a truth table as your input and ideal values?

Question

I have a truth table representing inputs and the results of running the XOR function on those inputs.

input1  input2   result
  0       0        0
  0       1